Alcohol Use Disorder (AUD) is substantially heritable syndrome with high public-health impact the etiology of which remains poorly understood. Over the last decade, genome-wide association studies (GWAS) have been successfully applied to an increasing number of complex biomedical and neuropsychiatric syndromes leading to important insights into biological causal pathways. Results have uniformly shown these disorders to be higher polygenic with small effect-size risk variants. However, progress in the molecular genetic dissection of AUD has been slow due to lack of adequate sample sizes of well-characterized severely ill cases. In response to this concern, NIAAA issued on 4/3/17 NOT-AA-17-002 ?Submission of Applications Containing Genome- Wide Association Studies.? This proposal was designed to meet the goals laid out by NIAAA in this announcement. Because of low recruitment costs, the viability of which have been tested by pilot studies, we can, within the confines of an R0-1 budget, assess and genotype 12,000 cases of severe DSM-5 AUD. However, this application should not be viewed as only a ?stand-alone? project, but as part of NIAAA?s plan to combine results across multiple studies funded under this initiative and ongoing efforts of the Psychiatric Genomics Consortium (PGC) to gain sufficient aggregate sample size for molecular analysis. Assessment of complete DSM-5 criteria for AUD, associated key co-morbidities (depression, drug abuse, antisocial personality) and other risk factors (e.g. personality) will be performed by an avatar aided tablet/web program. We will collect a severe and ethnically diverse sample from two clinical data collection networks (Hazelden- Betty Ford and Baltimore Treatment Network) and via the web from the Faces and Voices of Recovery. Scientific aims include i) identification, from public sources, of ethnically matched screened alcohol-exposed controls with GWAS, ii) complete imputation and quality control checks of genotype data, iii) performance of GWAS analyses of each super-population and a meta-analysis across ethnic samples, iv) establishment of collaborative links with the PGC and other NIAAA projects to implement cross-samples analyses, v) performance of gene and geneset analyses to seek further insights into risk pathways to AUD, vi) examination of phenotypic heterogeneity within AUD using our rich measures of symptomatology, personality, and comorbidity and attempted molecular validation of these subtypes and vii) use of polygenic risk scores (with our study serving as both a test and training sample) and genetic correlations to better understand the commonality of genetic risk between AUD and other key psychiatric and substance using disorders. With this and parallel similar projects, sufficient sample sizes of AUD cases and associated controls will be ascertained to identify risk SNPs, trace these to their relevant risk genes and use these aggregate findings to provide insight into biological etiologic pathways to AUD risk. Such pathways can then lead to new potential treatment and prevention strategies for the critical but poorly understood syndrome of AUD.